Ravi,
I wish I could agree re the common vocabulary being feasible,
but the evidence I get has consistently opposed that conjecture.
From discussing EHRs with doctors in private practice, I find
that they hate it. It does use a common medical vocabulary in EHR
specifications, which are basically a class of XML specs.
The docs hate putting in all that data which is so tightly
specified that the EHR costs them a lot of extra time. They try to get their MTs
to do it for them, which makes a real mess. They agree that they get more
revenue from the insurance companies with such a tight medical specification,
but they hate that they have to spend so much time away from the patient.
The profession is being forced to change so that the information
system can work better. I thought the priorities were the other way around.
I saw the same thing happen in a 911 response system case. The
system took 20 minutes to enter the data for a new customer, largely because
the fields in each form were so tightly type constrained that they were without
option even in emergencies; the insurance data had to be in before they could
get authorization to send out the EMTs and/or the police.
In the software engineering world over the last five decades,
the same process happened. It changed from a fun research area to a sweat shop
populated by software engineers, who came and went.
Physical therapists, year by year, are being required to get
more and more degrees and certificates or else new grads will not be able to
practice. The older practicing therapists are grandfathered in of course, or
they would have opposed it.
The answer is not tighter constraints on workers. Automate
them, but don't force each one to fit a canned role or they lose their
creativity and motivation.
Replace them completely if possible, but constraining them event
by event simply makes the activity unworkable sooner or later.
The short term business result is good - investors get better
returns in the short term by forcing 1984 like practices. But in the long run,
it goes the way of so many other businesses that automated their processes into
useful valuable commodities. Then they get replaced by new developments.
The common wisdom rule is that people improve performance by
twenty to thirty percent every time they double their experience. That rule is
commonly used in estimating progress on large projects involving learning. But
when there is nothing to learn, and the tasks are rigidly prescribed, it's time
for a computer.
Sincerely,
Rich Cooper,
Rich Cooper,
www DOT EnglishLogicKernel DOT com
Rich AT EnglishLogicKernel DOT com
( 9 4 9 ) 5 2 5-5 7 1 2
From: ontolog-forum-bounces@xxxxxxxxxxxxxxxx
[mailto:ontolog-forum-bounces@xxxxxxxxxxxxxxxx] On Behalf Of Ravi Sharma
Sent: Tuesday, March 17, 2015 3:28 PM
To: [ontolog-forum]
Subject: Re: [ontolog-forum] Endurantism and Perdurantism - Re: Some
Comments on Descriptive vs. Prescriptive Ontologies
Agree thanks, just one more thought - as it takes a lot of
effort to reconcile and align different meanings and different words, as we do
in databases, we need to start encompassing best practices for not having
multiple names for same ontology applications similar to master data.
Bioinformatics ontologies have accomplished this already to some extent? can
the synonym or such operations can be done online or near-line to avoid
duplicates?
On Tue, Mar 17, 2015 at 3:11 PM, Rich Cooper <Rich@xxxxxxxxxxxxxxxxxxxxxx>
wrote:
Dear
Ravi,
I
agree that they should share a common SUBVOCABULARY of names with SIMILAR
meanings to the participants, but not an entire Vocabulary shared by all
concerned.
For
example, the various internet protocols establish a hierarchy of component
protocols, and the full vocabulary of UDP up through HTTP, one level at a
time. My email client doesn't have to know the lower or upper
vocabularies, and there may be many email clients that implement that oasis in
the market spectrum in very different ways.
Again,
two communicants intercommunicating have to have some common words. But
there is no reason for all communicants to have the full oasis dialect, plus
all the neighboring dialects in the community. Also, that is less secure
than keeping each protocol compartmentalized.
Two
email programs taken at random provide very different solutions to the
conceptually simple email oasis. They share a functional market, but with
divergent customer needs. My needs for email are different from
yours. The best email programs are the ones that have enough variability
through the settings to appeal to many small markets, but there is no single
dominant email client.
Sincerely,
Rich
Cooper,
Rich
Cooper,
www
DOT EnglishLogicKernel DOT com
Rich
AT EnglishLogicKernel DOT com
( 9 4
9 ) 5 2 5-5 7 1 2
For
Rich - how will we share vocabularies or share information, knowledge or
data if we can not agree on vocabularies that are common to SME domains?
On
Tue, Mar 17, 2015 at 12:20 PM, Amanda Vizedom <amanda.vizedom@xxxxxxxxx>
wrote:
Rich,
Actually,
that bit you quoted is from Pat.
Dear
Amanda,
You
wrote:
Yes, but more than that. In addition, the actual logic allows each point of
view to be expressed in ways that 'natural' for it, and appropriate conclusions
drawn within that point of view, and yet both POVs can use the same vocabulary
and be not only mutually consistent (in the strict logical sense) but even
interderivable from one another, given appropriate linking axioms. So a
conclusion stated in one style can be interderivable with the same conclusion
stated in the other style.
While
that is true, it should also be that SMEs 1 and 2 would describe a slightly
different view, using slightly different vocabulary, not only in the orthogonal
sense of logic for different views, but in the sense of the reasoning each uses
to customize the individual history with that semantic belief of 1, versus that
semantic belief of 2.
In
every KE project I have worked on, conducted or been involved with, there are
subtle differences in world view of each participant. If they got to
freely discuss and debate wherever problems arose, it could get intense, but it
worked more smoothly than the sweat shop management method. If the
manager or customer determined a highly detailed work plan without discussing
it with KEs in the same detail, the plan would be very unstable, with lots of
unanticipated problems that vary by the week.
So
requiring a single POV's vocabulary to match another's simply hasn't worked
historically. Therefore, I expect each claim in a patent to have a unique
vocabulary beyond the claim construction vocabulary. So far, they do, and
by experience I know that the crucial issues will be decided over those words
and phrases that are not standard. Each claim represents one POV as
specified, edited, reviewed, and ultimately issued.
Sincerely,
Rich
Cooper,
Rich
Cooper,
www
DOT EnglishLogicKernel DOT com
Rich
AT EnglishLogicKernel DOT com
( 9 4
9 ) 5 2 5-5 7 1 2
Comment
below.
On
Tue, Mar 17, 2015 at 1:46 PM, Pat Hayes <phayes@xxxxxxx> wrote:
On Mar 17, 2015, at 2:24 AM, Matthew West <dr.matthew.west@xxxxxxxxx>
wrote:
> Dear Pat,
> It seems to me you are saying that the data elements that you need to talk
> about what is true at a time and change over time are the same for the
> different views of how the world is.
Yes, but more than that. In addition, the actual logic allows each point of
view to be expressed in ways that 'natural' for it, and appropriate conclusions
drawn within that point of view, and yet both POVs can use the same vocabulary
and be not only mutually consistent (in the strict logical sense) but even
interderivable from one another, given appropriate linking axioms. So a
conclusion stated in one style can be interderivable with the same conclusion
stated in the other style.
> I agree. If you are to provide an
> adequate description of however the world is from different view points, I
> think that must be so even. If not the different views would not work, and
> would be dismissed. The problem is that the different views are workable,
> rather than demonstrably wrong.
> However, I don't think that a data format that obfuscates the different
> views helps. It does not make the different views the same somehow, it
just
> demonstrates some level of equivalence. Equivalence is not the same as
> compatibility.
Indeed, but one gets actual compatibility from this style of axiomatization.
(Well to be very careful, there are things that can only be stated in one
framework, but nothing can be done about that.)
Pat
Yes,
and more than that: The approach Pat describes does not "obfuscate
the different views." Having sufficient expressiveness and logical features
allows one to represent very many things as seen from both + perspectives. If
you are working on a project for which it is useful to have that flexibility,
it is very, very likely that it will also be useful to *capture*, that is
*model* the different views more or less explicitly. You will also want not
only model-parts that let you convert between the two, but inference support
that actually does this, and easily.
Having
these things is very useful in enabling multi-modal, multi-directional, and
flexible interaction between an ontological knowledge base and (a) people, (b)
varied systems, and (c) complementary sub-systems such as NLP that have their
own internal representation and can be used more efficiently if they can
interact with the knowledge base's contents in a particular form. In other
words, such flexibility, with explicit (non-obfuscated) axiomitization of
various models and relationships between them, enables a wider variety of
applications, guis, and systems to work with the same ontology / knowledge
base, each interacting with the knowledge (semi-)normalized to the form most
conducive to that interaction.
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
--
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
--
Thanks.
Ravi
(Dr. Ravi Sharma)
313 204 1740 Mobile
|
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J (01)
|